As the national discussion about who should be president after 2020 is heating up, so is the conversation about how we should pick presidents. The reform movement known as the National Popular Vote (“NPV”) interstate compact plan has gained significant momentum; in the past month alone, three additional states—Colorado, Delaware and New Mexico—have joined the agreement, which would require each signatory state to select electors pledged to support not the candidate who garners the most popular votes in that state, but the candidate who wins the most votes nationally. By its terms, the compact will become effective only when states totaling at least 270 electoral college votes—the number needed to win the White House—enact the plan. The three recent pickups mean that 14 states (plus DC, which is not a state but which participates in the electoral college) have adopted; these 15 jurisdictions represent 189 electors—70% of the needed 270.
In 2001, in some academic writings, my brother (and fellow law professor) Akhil Amar and I—and, separately, another law professor, Robert Bennett—laid out the intellectual foundations of this plan that would move the country close to having a national popular election for president, even though the electoral college would still exist and do the actual presidential picking. At the time, few people expected states to act on these ideas. But, thanks in large part to an organization headed by Stanford mathematician John Koza, here we are, two decades later, moving forward.
As I’ve noted before, the path ahead for NPV gets tougher, though, because thus far the signatories have been Blue states, and Red state legislatures are skeptical of the idea, seeing it as a partisan play to help elect Democratic presidents. NPV cannot, and should not, succeed if people view it as a partisan trick. In fact, it is no such thing; in 2000 many analysts expected Republican George Bush to win the popular vote and lose the electoral college, and in 2004 Bush beat John Kerry soundly in the national vote tally but would have lost the electoral college vote and thus the White House had things been a bit different (60,000 voters switching sides) in Ohio. And more generally, as illustrated even in 2016, we cannot easily know which candidate would have won past presidential elections had the NPV plan had been in effect, because the candidates would have campaigned differently; for example, it is possible that Mr. Trump could have run up the margin of victory in some solidly Red states, and narrowed the margin of loss in some deep Blue ones.
But Republican stalwarts—especially elected Republican legislators—remain worried that NPV is simply a tactic to elect folks like Bernie Sanders. And in 2019 America, politics often overtakes analytics. So here’s an idea: Let’s take ourselves out of 2019 America, and transport ourselves 12 years—three presidential election cycles—ahead. How could we do that? Simple. Commit to adopting the NPV concept, but delay its implementation until a time far enough in the future so that neither side could know what the partisan consequences might be. If Red state voters and elected officials believe (as do I) that all votes should count equally nationwide and also that presidential elections should not focus only on a handful of swing states and ignore states like Texas and California that are solidly in the R or D column from the outset, they could simply agree to adopt NPV now, but declare that their adoption won’t take effect until the election of 2032. Nobody today can reasonably predict which party might be helped or hurt by a national vote at that time. And because Red State adoption of NPV is needed to get over the 270 threshold, any Red state that adopts the measure on these terms can feel confident that its 2032 timeline would be respected. (If Red states fail to join, we can know that their concern is not fear of a Democratic power play but something else entirely.)
This focus on the present and the future also directs attention where it should be, and underscores how the debate (which has heated up in recent weeks on some op-ed pages) over how much the electoral college was originally intended in 1787 to protect slavery is largely beside the point. I think the historical evidence strongly suggests that placating Southern states and the institution of slavery was a primary driver of electoral college design, but even if that were not the case, the modern arguments against the electoral college should carry the day anyway. An innocent original intent does not cure the problems with the electoral college in modern American democracy, any more than it cures many other innocent defects in the original design (such as indirect election of senators) that have been undone by formal or informal amendment. In this regard, it is telling that no state would ever think of electing governors via an intrastate electoral college even if such an approach weren’t unconstitutional under the one voter, one vote principle that applies outside of Senate and presidential elections. The most plausible and legitimate reason not to move in the direction of NPV is not the innocent underpinnings of the electoral college; it is the fear of unintended and unforeseen consequences. The electoral college, while far from perfect, has served the nation reasonably well, and change always brings the risk of glitches.
But here too, the deferred implementation approach I advocate for gives all the signatory states—and also, importantly, Congress—time to smooth out some of the rough spots in the current NPV compact details. In particular, substantial nonuniformity among the states in the national vote count on the questions of who votes, how the votes are cast, and how they are counted and recounted undermines the normative appeal of the move to a national popular vote, and it raises the specter of electoral crises that can and should be avoided. My brother and I noted—indeed highlighted—this problem when we first analyzed the idea of state-level nonconstitutional movement towards direct election almost two decades ago. Here is what we wrote:
Of course, any coordinated state-law effort would require specifying key issues: Majority rule or plurality rule? Runoff or no? How should recounts and challenges be handled?
It would be hard to rely completely on the laws and courts of each state, many of which might not be part of the cooperating 270 group. . . . What if some state let [seventeen-year-olds] vote in an effort to count for more than its fair share of the national total? And what about Americans who live abroad or in the federal territories?
These questions suggest an even more mind-boggling prospect: our national-vote system need not piggyback on the laws and machinery of noncooperating states at all! Let these noncooperating states hold their own elections, but so long as they amount to less than 270 electors, these elections would be sideshows. The cooperating states could define their own rules for a uniform “National Presidential Vote” system. In that case, our law would read something like this:
Section 1. This state shall choose a slate of electors loyal to the Presidential candidate who wins the “National Presidential Vote,” if and only if other states, whose electors taken together with this state‘s electors total at least 270, also enact laws guaranteeing that they will choose electors loyal to the Presidential candidate who wins the “National Presidential Vote.”
Section 2. The “National Presidential Vote” shall be administered as follows. . . .
Section 2 of this model law would proceed to specify the precise rules of this “National Presidential Vote.” For example, Section 2 could provide that Americans everywhere who want to be counted must register in a system to be administered by a nongovernmental election commission . . . . Section 2 could also specify uniform rules of voting eligibility, uniform presidential ballots, and an election dispute procedure . . . . Alternatively, Section 2 might contemplate that the “National Presidential Vote” should be administered by a new interstate election council or directly by the federal government; and Congress could then pass a statute blessing this more elaborate interstate agreement.
The drafters of the NPV plan in play today did not take our advice in this respect; they did not build into the plan uniform rules of voting eligibility, uniform presidential ballots, and an election dispute procedure. Nor did they delegate authority to do so to a nongovernmental commission. Nor did they affirmatively invite Congress to step in. To my mind, the most important issue surrounding the entire NPV movement as it approaches critical mass in the states and builds some support on Capitol Hill—and the issue on which thoughtful analysts ought to be focused—is how Congress, in the course of approving the Compact or in adopting a freestanding statute, might fill in the dangerous gaps in the NPV design. If and when the NPV comes into being, I would forcefully urge Congress to supplement it with a system of uniform rules for tallying sentiment in all fifty states.
In our earlier writings, my brother and I suggested that it might be possible that the NPV idea could be implemented without congressional approval as an interstate compact. But as I explain in more depth in some law review articles: (1) the failure of the NPV folks to specify and implement any uniform voting and counting procedures; (2) the involvement of the District of Columbia—which is under Congress‘s jurisdiction and whose involvement without congressional approval perhaps should not count towards the magic 270 number and; (3) the helpfulness of Congress and the Supremacy Clause in reinforcing any prohibition on late withdrawal by states, combined with murkiness in Supreme Court cases about precisely what constitutes a Compact between states that requires approval, all lead me today to argue strongly in favor of congressional action (which I have argued Congress has the power to take) to supplement any improvements in the NPV design that signatory states can agree upon during the period prior to implementation.